Ultra-high resolution photon-counting coronary CT angiography improves coronary stenosis quantification over a wide range of heart rates – A dynamic phantom study

Abstract

PurposeTo investigate the effect of using photon-counting detector (PCD)-CT with ultra-high resolution (UHR) on stenosis quantification accuracy and blooming artifacts from low to high heart rates in a dynamic motion phantom. MethodTwo vessel phantoms (diameter: 4 mm) containing solid calcified lesions (25%, 50% stenoses), filled with different concentrations of iodine, inside an anthropomorphic thorax phantom attached to a coronary motion simulator were used. Scanning was performed on a PCD-CT system using an ECG-gated mode at UHR and standard resolution (SR) (0.2, 0.6 mm slice thickness, respectively). Images were reconstructed at 60, 80 and 100 beats per minute (bpm) (UHR: Bv56 kernel, quantum iterative reconstruction (QIR) level 3; SR: 55 keV, Bv40 kernel, QIR3). Percent diameter stenosis (PDS) and blooming artifacts were measured by two readers. ResultsPDS measurements derived from UHR were more accurate than SR for both lesions at every heart rate (p ≤ 0.005 for all, e.g. 50% lesion SR vs. UHR: at 60 bpm 57.1% [55.2–59.2] vs. 50.0% [48.5–51.2], at 100 bpm 61.0% [58.6–64.3] vs. 52.4% [51.3–54.3]). Overall mean difference across heart rates and lesions compared to the nominal stenoses was 9.2% (Limit of Agreement (LoA), 2.4%/16.0%) for SR vs. 2.4% (LoA, −2.8%/7.5%) for UHR. Blooming artifacts decreased with UHR compared to SR for both lesions at every heart rate (p < 0.001 for all, e.g. 50% lesion SR vs. UHR: at 60 bpm 63.8% [60.6–69.5] vs. 52.5% [50.0–57.5], at 100 bpm 70.2% [64.8–78.1] vs. 56.1% [51.2–60.8]). ConclusionsThis motion phantom study demonstrates improved stenosis quantification accuracy and reduced blooming artifacts with UHR-PCD-CT compared to SR, independent of heart rate.